| { | |
| "best_metric": 0.7282186150550842, | |
| "best_model_checkpoint": "mobilebert_add_GLUE_Experiment_logit_kd_sst2_128/checkpoint-2635", | |
| "epoch": 10.0, | |
| "global_step": 5270, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.9e-05, | |
| "loss": 1.5487, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5779816513761468, | |
| "eval_loss": 1.3928688764572144, | |
| "eval_runtime": 1.4162, | |
| "eval_samples_per_second": 615.727, | |
| "eval_steps_per_second": 4.943, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.3629, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5504587155963303, | |
| "eval_loss": 1.4979432821273804, | |
| "eval_runtime": 1.4117, | |
| "eval_samples_per_second": 617.704, | |
| "eval_steps_per_second": 4.959, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.7e-05, | |
| "loss": 1.1397, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6754587155963303, | |
| "eval_loss": 1.3926728963851929, | |
| "eval_runtime": 1.4106, | |
| "eval_samples_per_second": 618.195, | |
| "eval_steps_per_second": 4.963, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.5649, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8073394495412844, | |
| "eval_loss": 0.7289367318153381, | |
| "eval_runtime": 1.4117, | |
| "eval_samples_per_second": 617.688, | |
| "eval_steps_per_second": 4.959, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4112, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8073394495412844, | |
| "eval_loss": 0.7282186150550842, | |
| "eval_runtime": 1.5054, | |
| "eval_samples_per_second": 579.245, | |
| "eval_steps_per_second": 4.65, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.3462, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.805045871559633, | |
| "eval_loss": 0.7653937339782715, | |
| "eval_runtime": 1.4197, | |
| "eval_samples_per_second": 614.218, | |
| "eval_steps_per_second": 4.931, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.3069, | |
| "step": 3689 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7970183486238532, | |
| "eval_loss": 0.8302631974220276, | |
| "eval_runtime": 1.4135, | |
| "eval_samples_per_second": 616.903, | |
| "eval_steps_per_second": 4.952, | |
| "step": 3689 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.2833, | |
| "step": 4216 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7924311926605505, | |
| "eval_loss": 0.880594789981842, | |
| "eval_runtime": 1.4122, | |
| "eval_samples_per_second": 617.467, | |
| "eval_steps_per_second": 4.957, | |
| "step": 4216 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 4.1e-05, | |
| "loss": 0.2662, | |
| "step": 4743 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7958715596330275, | |
| "eval_loss": 0.9296879768371582, | |
| "eval_runtime": 1.4088, | |
| "eval_samples_per_second": 618.966, | |
| "eval_steps_per_second": 4.969, | |
| "step": 4743 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 4e-05, | |
| "loss": 0.2521, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7717889908256881, | |
| "eval_loss": 1.097933292388916, | |
| "eval_runtime": 1.4221, | |
| "eval_samples_per_second": 613.19, | |
| "eval_steps_per_second": 4.922, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 5270, | |
| "total_flos": 1.607798324461568e+16, | |
| "train_loss": 0.6482070778081268, | |
| "train_runtime": 2637.7784, | |
| "train_samples_per_second": 1276.624, | |
| "train_steps_per_second": 9.989 | |
| } | |
| ], | |
| "max_steps": 26350, | |
| "num_train_epochs": 50, | |
| "total_flos": 1.607798324461568e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |